U.S. patent number 8,260,109 [Application Number 11/492,681] was granted by the patent office on 2012-09-04 for system for digital time shifting and method thereof.
This patent grant is currently assigned to ATI Technologies ULC. Invention is credited to Branko D. Kovacevic.
United States Patent |
8,260,109 |
Kovacevic |
September 4, 2012 |
System for digital time shifting and method thereof
Abstract
A multiplexed packetized data stream carrying real-time
multimedia programs is received at a first hardware demultiplexer.
Based on a user input, a video and timing portion of a program
associated with the multiplexed packetized data stream can be
stored for subsequent display. One type of subsequent display is
time shifted display, where the stored portion of the program is
played back while new portions of the program are being stored.
During time shifted play back, a second hardware demultiplexer can
be used, so that one demultiplexer stores new data and maintains a
current clock value while the other decodes and displays the stored
data.
Inventors: |
Kovacevic; Branko D. (Toronto,
CA) |
Assignee: |
ATI Technologies ULC (Markham,
Ontario, CA)
|
Family
ID: |
36821788 |
Appl.
No.: |
11/492,681 |
Filed: |
July 25, 2006 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20070009236 A1 |
Jan 11, 2007 |
|
Current U.S.
Class: |
386/239 |
Current CPC
Class: |
H04N
5/76 (20130101); H04N 9/8042 (20130101); H04N
9/8205 (20130101); H04N 2005/91328 (20130101) |
Current International
Class: |
H04N
9/80 (20060101) |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
0942603 |
|
Sep 1999 |
|
EP |
|
1014730 |
|
Jun 2000 |
|
EP |
|
Other References
"Transmission of Non-Telephone Signals," Information
Technology--Generic Coding of Moving Pictures and Associated Audio
Information: Systems, ITU-T Recommendation H.222.0, Jul. 1995, 120
pages. cited by other .
"Information Technology--Generic Coding of Moving Pictures and
Associated Audio Information--Part 3: Audio," ISO/IEC 13818-3,
Second Edition, Apr. 15, 1998, 116 pages. cited by other .
Chen, Zhigang et al.,"Real Time Video and Audio in the World Wide
Web," 17 pp. [online] [Retrieved from the Internet Apr. 2, 2008 at
URL
<http://www.geckil.com/.about.harvest/mirror/www.w3.org/Conferences/WW-
W4/Papers/p211/>. cited by other .
Civanlar et al., "RTP Payload Format for Bundled MPEG," May 1998, 8
pp. [online] [Retrieved from the Internet Apr. 2, 2008 URL
<http://www.geckil.com/.about.harvest/rfc/rfc2343.txt>. cited
by other .
Hoffman et al., "RTP Payload Format for MPEG1/MPEG2 Video," Jan.
1998, 15 pages, [online] [Retrieved from the Internet Apr. 2, 2008
URL <http://www.geckil.com/.about.harvest/rfc/rfc2250.txt>.
cited by other .
Bolot et al., "Adaptive Error Control for Packet Video in the
Internet," International Conference on Image processing, 1996.
Proceedings, pp. 25-28, vol. 1, Sep. 16-19, 1996, ISBN
0-7803-3259-8. cited by other .
Feng, Jian et al., "Error Concealment for MPEG Video
Transmissions," Mar. 13, 1997, IEEE Transactions on Consumer
Electronics, pp. 183-187, vol. 43, Issue 2, ISSN 0098-3063. cited
by other .
Le Gall, Didier "MPEG: Video Compression Standard for Multimedia
Applications," Communications of the ACM, Apr. 1991, vol. 34, No.
4, pp. 47-58. cited by other .
Zhu, Wenw et al., "End-to-end Modeling and Simulation of MPEG-2
Transport Streams over ATM Networks with Jitter," 1998 IEEE
Transactions on Circuits and Systems for Video Technology, vol. 8,
No. 1, Feb. 1998, pp. 9-12.ZHU. cited by other .
Kassler et al., "Simulating MPEG-2 Transport Stream Transmission
over Wireless ATM," 1999 IEEE, 0-7803-5041-3/99, pp. 3057-3060.
cited by other.
|
Primary Examiner: Vaughn, Jr.; William C
Assistant Examiner: Tekle; Daniel
Claims
What is claimed is:
1. A method comprising: receiving a multiplexed packetized data
stream that carries real-time multimedia programs; during a first
time: parsing packet identifiers of the packetized data stream to
identify video data and timing data associated with a first
multimedia program, the timing data identified based on timing
information transmitted in packets of the multiplexed packetized
data stream; storing, in response to parsing during the first time,
a first portion of the first multimedia program, the first portion
comprising first video data associated with a first packet
identifier and first timing data associated with a second packet
identifier ; and setting a system time indicator to a stored system
time value, wherein the stored system time value is based on a
portion of the timing data of the first portion of the packetized
data stream; during a second time: incrementing the system time
indicator; retrieving the first video data of the first portion of
the packetized data stream for video decoding; and parsing packet
identifiers of the packetized data stream to identify video data
and timing data associated with the first multimedia program;
storing, in response to parsing during the second time, a second
portion of the packetized data stream, the second portion
comprising second video data associated with the first multimedia
program and second timing data of the first multimedia program.
2. The method of claim 1, wherein storing includes providing the
first video data to a first ring buffer associated with the first
packet identifier.
3. The method of claim 2, wherein storing further includes
providing the first video data from the first ring buffer to a hard
disk.
4. The method of claim 3, wherein parsing includes parsing a
transport stream header of a packet at an integrated circuit device
while the packet associated with the transport stream header is
being received at the integrated circuit device.
5. The method of claim 4, wherein storing further includes
controlling access of the first ring buffer at the integrated
circuit.
6. The method of claim 5, wherein storing includes providing the
first timing data to a second ring buffer associated with the
second packet identifier.
7. The method of claim 6, wherein storing further includes
providing the first timing data from the second ring buffer to the
hard disk.
8. A system comprising: a first input to receive a multiplexed
packetized data stream that carries real-time multimedia programs;
a first transport stream demultiplexer having an input coupled to
the first input to select packets of data having a predefined
packet identifier and an output to provide the select packets of
data; a storage device having a data port coupled to the output of
the first transport stream demultiplexer to receive the select
packets, wherein the storage device is to store the select packets;
a first clock recovery module having an input coupled to the first
input, and an output, wherein the first clock recovery module is to
generate a clock at the output based upon received timing
information transmitted in packets of the multiplexed packetized
data stream, the first clock recovery module to set a system time
indicator based on the received timing information and to increment
the system time indicator to generate the clock; and a decoder
having a first input coupled to the output of the first clock
recovery module to receive the clock, a second input coupled the
data port of the storage device to receive the select packets, and
an output to provide decoded real-time data.
9. The system of claim 8, wherein the first transport stream
demultiplexer is at an integrated circuit and comprises a first
parser to parse a header of transport stream while the transport
packet is being received.
10. The system of claim 9, wherein the first transport stream
further comprises a second parser to parse a header of packetized
elementary stream (PES) stream.
11. The system of claim 10, wherein the second parser is further to
parse the header of packetized elementary stream (PES) stream
within a transport stream packet payload.
12. The system of claim 8 wherein the first clock recovery module
is at the integrated circuit.
13. The system of claim 8 wherein the first transport stream
demultiplexer is at an integrated circuit and comprises a parser to
parse a header of packetized elementary stream (PES) packet within
a payload of a transport stream packet being received.
Description
CROSS-REFERENCE TO RELATED APPLICATION
The present application claims priority from U.S. patent
application Ser. No. 09/707,060, filed Nov. 6, 2000, entitled
"SYSTEM FOR DIGITAL TIME SHIFTING AND METHOD THEREOF," naming
inventor Branko Kovacevic, which application is incorporated by
reference herein in its entirety.
FIELD OF THE DISCLOSURE
The present invention relates generally to time shifting of video
data, and more specifically to time shifting of digital video
data.
BACKGROUND
Systems for time shifting a viewed program are known in the
industry. For example, if a viewer is interrupted by a phone call
during a television program, the program can be recorded for a few
minutes and then played back from the point of interruption while
addition video information is continually recorded. One prior art
method of accomplishing time shifting is to capture the rendered
video signal. When the rendered signal is an analog signal it is
digitized and stored. When the rendered signal is a digital signal
it can be captured directly. Once captured, the rendered digital
data can be stored directly. A digital signal stored directly can
require a large amount of storage space, even when only a few
minutes of video are captured. The digital signal can be compressed
to reduce the amount of storage space required. However,
compressing a video signal requires additional processing power,
resulting in additional costs.
As the use of digital video data becomes increasingly common, a
method and apparatus for time shifting a digital program that is
more efficient than those known in art would be advantageous. One
known method to provide digital video data is to provide the data
using a specific protocol that has the ability to transmit the
digital video data in a compressed format. An example of one such
format is known as MPEG-2, and has been approved by the
International Organization for Standards (ISO) Moving Pictures
Experts Group (MPEG group). MPEG-2 is a versatile communication
standard that gives theoretical explanations needed to implement an
MPEG-2 decoder through the syntax and semantics of coded
bit-streams. MPEG-2 is an open standard and continues to evolve and
be applied to a wide variety of applications ranging from video
conferencing to High Definition Television (HDTV). The MPEG-2
standard, as a generic and open standard, is intended for variety
of audio/video coding applications.
One method of transporting large amounts of various types of
transport stream data is to use a multiplexed packetized data
stream capable of carrying real-time multimedia programs. One
example of a multiplexed packetized data stream is described in the
standard ISO/IEC 13818-1 and will be referred to as a transport
stream. Transport streams generally offer robustness for noisy
channels and can carry multiple programs (like multiple TV
services) within the same multiplex. The transport stream is based
on 188 byte long packets that are well suited for hardware error
correction and processing schemes needed in noisy environments,
such as coaxial cable television networks and satellite
transponders. Such a transport stream facilitates fast program
access, channel hopping and synchronization between multiple
programs within the transport stream.
A transport stream consists of fixed length packets based on 4
bytes of header followed by 184 bytes of data payload, where data
payload is obtained by partitioning larger data blocks. For
example, an elementary stream (ES) is a set of data generally
consisting of compressed data from a single source, such as a video
or audio source, with some additional ancillary data for
identification, characterization and synchronization. ES streams
are first packetized into either constant length or variable length
Packetized Elementary Stream packets (PES packets) consisting of a
header and payload. Each PES packet header starts with start code
(ox000001) followed with the stream_id byte identifying type of ES
underneath.
PES packets from various elementary streams are merged together to
form a program (service) with its own system time clock (STC). All
ES component streams within one program are synchronized have
periodic PTS stamps corresponding to the STC counter to indicate
the proper timing for each ES.
The relatively long and most often variable length PES packets are
further packetized into shorter TS packets having a constant size
of 188 bytes. A small and constant TS packet size makes error
recovery easier and faster. Usually, the transport stream carries
several programs, each with its own STC. Each TS packet consists of
a TS Packet header with optional Adaptation Field followed by
useful data payload containing portion of a PES packet. The TS
header consists of a sync byte, flags, indicators information for
error detection and timing and Packet_ID (PID) field used to
identify elementary stream carried underneath of a PES packet. In
addition to identifying specific elementary streams, one PID is
used to identify a program specific Information (PSI) table
data.
Each TS PSI table is sent in sections, usually occupying one or
more TS packets. Four types of PSI tables exist: 1) Program
Association Table (PAT) listing unique program_number (as an
identifier of each program in one multiplex) and PID of the PMT
table; 2) Program Map Table (PMT) listing PIDs of all component
streams making a given program. PMT may be constructed for each
program separately or be common for a group of programs; 3)
Conditional Access Table (CAT) identifying PID of Entitlement
Management Messages and ID of used conditional access system if any
scrambling of TS or PES packets is done; 4) Private Table carrying
Network Information Table (NIT) or private data.
The Hierarchical structure which exists between ES streams, PES and
TP packets is illustrated in prior art FIG. 1-4.
A method and apparatus for efficient time shifting of multiplexed
packetized data streams, such as a packet stream, would be
advantageous.
BRIEF DESCRIPTION OF THE DRAWINGS
FIGS. 1-4 illustrate various information associated with an MPEG
transport stream of the prior art.
FIG. 5 illustrates in graphical form a time line indicating various
modes of operation in accordance with the present invention;
FIG. 6 illustrates in block diagram form a specific embodiment of a
system having to digital transport stream receivers in accordance
with the present invention.
DETAILED DESCRIPTION OF THE DRAWINGS
A specific method and apparatus is disclosed describing a time
shifting technique. In one embodiment, the disclosed time shifting
technique can be based upon a hardware transport stream
demultiplexer that interfaces to a transport stream. The hardware
demultiplexer application assists in the extraction and parsing of
a multiplexed packetized data stream, such as a MPEG-2 Transport
Stream (TS) multiplex. One such hardware demultiplexer is disclosed
in pending U.S. patent application Ser. No. 09/489,682, entitled
"METHOD AND SYSTEM FOR ACCESSING PACKETIZED ELEMENTARY STREAM DATA"
(0100.9901350), which is hereby incorporated herein by reference.
The disclosed hardware transport core is used to filter component
streams into 15 memory ring buffers, one allocated in the frame
memory for the dedicated MPEG-2 video decoder and others in the
system memory for the dedicated software parser. It can demultiplex
the most frequent transport packets of video stream into an
Elementary Stream (ES) by monitoring the first packet identifier
(PID) of each TS packet. This flexible filter can be set to extract
private data from the adaptation field (AF) or from the PES packet
header. Thirty-one other PIDs can be simply filtered and routed to
a common (joint) or individual memory buffers for subsequent
software processing on the host processor. The basic idea of a time
shifting is shown in FIG. 5.
FIG. 5 illustrates three functions performed by a time shifting
system. A first function is to receive a live broadcast stream 510.
According to the graph of FIG. 5, the live broadcast stream is
continuously received during the time represented in FIG. 5.
A second function of a time shifting system is to record a specific
program after a user activates the time shifting feature. Vector
520 of FIG. 5 indicates when a specific program is being recorded
by the time shifting system.
A third function of the time shifting system is to display the
specific program. Vector 530 of FIG. 5 indicates when a specific
program is being played back. Specifically, vector portion 531
represents the time where the program is being displayed directly
from the live broadcast stream. Vector portion 532 represents the
time that the user is unable to view the program, i.e. the user is
away from the television. Therefore, in one embodiment, during this
time no program is displayed. In an alternate embodiment, the live
feed can continue to be displayed, even though the program is being
recorded.
Vector 533 represents the time during which the time-shifted
program, which has been stored, is being replayed at a normal
playback rate. Note that during this time, the live program feed
continues to be recorded for future time shifted play back.
Vector 534 represents a time during which the time shifted program
is being replayed at a faster than normal replay rate. By being
able to playback at a faster than normal rate, it is possible to
catch-up to the live broadcast stream.
The receive-only mode of vector 31 represents where the digital
transport stream receiver (DTSR) is receiving a live broadcast and
demultiplexing one program of a plurality of programs available in
the live broadcast stream. This will be referred to as Transparent
Mode indicating the transport stream is accessed immediately and
not saved. Therefore, from the point of view of digital storage
media (DSM), the received data is transparent.
Note that the PAT table is constantly acquired, in transparent
mode, and other modes, so that version number change or PMT table
PID change for a currently viewed program can be detected. If such
a change occurs during the live broadcast of a program, PIDs will
be reprogrammed for video and splicing with be handled.
A Continuous Time Shifting Mode occurs during vectors 532-534.
Continuous time shifting mode occurs when time shifting is selected
by the viewer to store part or all of a program for later viewing
after a short or long intermission. During continuous time shifting
mode, a selected program from a given multiplex is received and
stored on a hard disk, or other storage media, in the form of full
transport stream packets or PES packets.
A Part-Time Time-Shifting Mode, when selected by the viewer, allows
for replay of a time shifted program or fast forward (FF) replay of
a time shifted program at user defined FF speed. In FIG. 5 this is
represented as vectors 533 and 534. In a specific embodiment
discussed herein, this time-shifting mode is the most demanding
mode of the 3 described modes because: the host CPU system is
receiving and storing a real time event; at the same time, the host
CPU is retrieving saved stream data from the disk; simultaneously
with first two operations, the host CPU is performing transport
stream de-multiplexing of video, audio, private and PSI/SI data on
a host CPU; and at the same time the host CPU is restoring PCR/PTS
time-base information as described later.
For some digital television applications, time-shifting may be
considered a peak event that occurs sometimes or occasionally.
However, some users may depend on it all the time, up to the end of
the current program once it was started. For those users, typical
operating state of the system is time shifting, de-coupled from the
live stream. Time shifting of the digital transport stream should
offer the same quality as from the live broadcast (source
stream).
Systems suitable for time-shifting need to simultaneously receive
and decode a transport stream and handle incoming source stream (to
process all PSI and SI data) and record incoming source stream as a
full entity or just its one program. Time shifting allows the
viewer to step away from the TV monitor without missing any of the
program parts. One embodiment of time shifting includes storing all
transport packets received on the transport stream. Another
embodiment of time shifting that is more efficient includes: 1)
selectingjust the transport packets of interest (PSI, SI, video,
audio and data packets) that constitute one program event to
minimize the bit-rate of the recorded stream, to minimize the
bandwidth through the host bus interface unit, and to minimize hard
disk head movement (if any); 2) increasing the amount of storage
and useful life of the hard disk; and 3) assuring that the amount
of data that needs to be processed by the host processor is
received and stored as: transport stream packets; PES packets of
video, audio, data, PSI and SI content, de-multiplexed transport;
or PES packets of video and audio and bus master compressed video
into the video bit-stream buffer of the MPEG video decoding
device.
Selection of just one time shifted program reduces the potentially
high bit-rate of a transport stream multiplex to a manageable size,
suitable for storage on current 10 GB hard disk units (two hours of
10 Mbps stream). Obviously, a large disk drive is needed to allow
any reasonable length of time shifting. In time shifting mode where
time shifted material is simultaneously received and stored, the
bit-rate of the host bus-interface unit (HBIU) needs to be double a
system where the HBIU is only responsible for playing a single
program stream. Generally the bandwidth needed is calculated to be
approximately 20 Mbps instead 10 Mbps.
Because closed or proprietary systems, such as set-top boxes,
usually do not share the hard disk drive with other systems, very
specialized disk drives for audio-video applications with
specialized interfaces can be used. Hard drive features that would
be advantageous include: 1) Increasing access speeds and sustained
sequence transfers in two directions; 2) Having deferred
re-calibration of drive heads to prevent glitches or latencies
during playback; 3) Having head offsets to prevent losing a
revolution when going from side to side on a platter; 4) Supporting
on the fly error correction; and 5) Having embedded multi-disk
drive units that decrease access latencies.
The operating system can play a significant role in the efficient
use of the drive by accessing most frequent video data in large
blocks and decreasing seek time. Generally, larger read/write
blocks increase efficiency of data storage and retrieval. Sometimes
they can cause unwanted glitches by increasing latency during
access.
The first time shifting mode of operation is a receive-only mode.
During receive-only mode of operation a master digital time
shifting receiver (DTSR) 610, of FIG. 6, is programmed to receive
and parse transport stream packets matching video and PCR PIDs. A
host CPU 632 is assisting MPEG-2 clock recovery, and the same
recovered clock data is supplying Master DTSR 610 and the Secondary
DTSR 620. In one embodiment, the recovered clock is provided to the
secondary DTSR 620 registers through the use of the system memory
controller 630. Also, the Master DTSR 610 is programmed to perform
PID filtering of audio, private, and PSI/SI PIDs programmed in the
auxiliary PID registers. Secondary DTSR 620 is programmed for PID
filtering operations on Video PID programmed on a first auxiliary
PID register. However, since the receiver is in receive-only mode,
the video transport packets in the ring buffer 624 are disregarded.
The clock recovery algorithm is suppressed on the secondary DTSR
620. Only STC of the slave DTSR is set upon the channel change.
Host CPU 632 performs PES parsing of audio transport stream
packets, decode and presentation of audio frames (on AC-97 codec or
wave device), and continuous parsing and data processing of PSI
sections monitoring real-time events like PID change, PCR
discontinuity or splicing of audio stream. This activity by the
host CPU 632 is part of the normal receive only mode of operation
where a specified channel is being decoded and displayed. Specific
systems and methods for supporting these processes are described in
the patent application already incorporated by reference.
When in continuous time-shifting mode of operation, the host CPU
632 performs additional processing including: retrieval;
multiplexing; time base corrections; storage of video audio,
private and PSI/SI transport stream packets from multiple buffers
614 allocated in the memory space of the host CPU. In one
embodiment, however, the master DTSR 620 is used to decode and
display video stream as describe previously with reference to
receive only mode. Transport packets from a common program are
retrieved from the buffer 614 and provided to a digital storage
media circular file system in a multiplexed manner. Multiplexing is
performed by inserting audio, video, private, and PSI/SI transport
stream packets to satisfy a group of relevant criteria.
Fundamental functions performed during continuous digital time
shifting include: 1) Preserving of original ES_rate of each
component stream; 2) Limiting PCR jitter of newly created single
program multiplex; 3) Preserving VBV_delay value (the number of
periods of a 90 KHz clock derived from the 27 MHz system clock that
the VBV shall wait after receiving the final byte of the picture
start code before decoding the picture) to insure non-interrupted
MPEG video decode after initial VBV_delay time in constant bit-rate
(CBR) stream environments; 4) Preventing underflow or overflow of
elementary stream decoder buffers in accordance with the T_STD
model defined in ISO/IEC 13818-1 standard; 5) providing PID values
in the video or audio TS packets that were originally defined in
the PMT section to be a video or audio PIDs. Alternatively, a new
artificial PCR stream can be separately created and injected as TS
PCR packets at the rate of at least 10 times per second to create a
new time base for decimated, time-shifted stream stored on the DSM.
Whereby, the original PAT transport packet is modified or a new PAT
packet is inserted into the stream instead of the original PAT
section to indicate a single program only whose PMT section
indicates video, audio, PCR and other PID that carry subtitles,
program descriptions, etc. As a stable clock source, STC of the
Master DTSR is used to measure elapsed time between two PCR
samples; 6) Providing PTS values in the video, audio or private
data streams by using STC of the Master DTSR as elapsed time
counter; and 7) Initializing STC of the playback DTSR device to a
first available PCR value encoded in the stream saved on DSM media,
immediately after channel change.
While in part-time digital time-shifting mode, the host CPU 632
performs some additional processing like retrieval and
de-multiplexing of the single program transport stream created in
continuous time digital time-shifting mode during a storage
process. Generally, the playback of the stored program is combined
with continued transport stream de-multiplexing and recording of
the real-time transport stream. Such a mode of operation is the
most intensive mode of operation because the host CPU 632 must
create/store a multiplexed single program transport stream from a
continued reception of a live broadcast; and retrieve and
de-multiplex saved content from a digital storage media while
performing transport stream de-multiplexing, audio decode, and bus
mastering elementary stream video to the MPEG video decoder.
In one implementation, an MPEG decoder associated with the Master
DTSR 610 is used to decode and display a video stream from a DSM
media and receive private data, and PSI/SI sections from a live
broadcast. In such a case, a video PID of the Master DTSR 610 is
disabled, while video data with its PTS information is fed directly
to the MPEG decoder using the system memory controller 630.
However, PCR PID is programmed on a Master DTSR so that MPEG clock
recovery continues from a live transport stream feed and is
supplied to the STC counters of both the master DTSR 610 and the
second DTSR 620. In one implementation, only the video PID is
programmed into the Slave DTSR for retrieving live video stream and
sending it to circular buffer on the host system in the form of a
full MPEG-2 transport stream packets, while the Master DTSR is used
to buffer the non-video components of a specific program.
In another embodiment, a different partition of the software tasks
is possible on the host CPU 632 to achieve all three modes of a
digital time shifting. In the second embodiment, a first DTSR is
used as a combo video-PCR only device, either to receive and decode
video from a live broadcast or from a DSM media. The PCR PID of the
first DTSR is programmed always to match live broadcast, and full
clock recovery is done by the first DTSR. A second DTSR can be used
in all 3 modes to receive video, audio, private data and PSI/SI
sections, all utilizing auxiliary PID filters and received as full
MPEG-2 transport packets arriving in the single memory queue. This
way, the temporal order of a stream and validity of the T-STD
decoder model is inherently preserved. Also, the amount of the host
DRAM memory required for queue allocation is less than in the first
case. In both embodiments, a quality digital stream time shifting
at the transport packet level is achieved.
In yet another operating mode, a different partition of the
software tasks is possible on the host CPU 632 to achieve all three
modes of digital time shifting by storing PES layers as a basic
format of the audio/video data saved on a DSM. In PES operating
mode, two hardware embodiments are possible, the same as in TP
operating mode.
In a first hardware embodiment, the first DTSM is used as a combo
device, to achieve playback of live or stored MPEG video and
reception of audio, private & PSI/SI content. The second device
is used only to receive and de-multiplex MPEG-2 video transport
stream and retrieve MPEG-2 elementary stream from a live broadcast.
Upon retrieval of ES video, PES packets are formed and stored on
the DSM media.
In the second hardware embodiment, the first DTSM is used as a
combo video-PCR only device, either to receive and decode video
from a live broadcast or from a DSM media. The PCR PID is
programmed always to match live broadcast, and full clock recovery
is done by the first DTSR. A second DTSR is used in all 3
time-shifting modes to receive audio, private data, PSI/SI
sections, by utilizing auxiliary PID filters to store the transport
packets to a single memory queue. That way, a temporal order of a
stream and validity of T-STD decoder model is already
preserved.
In yet another time shifting embodiment, the video is
de-multiplexed to the level of elementary stream and stored at the
bit-stream buffer of the MPEG video decoder physically allocated in
the frame memory. The MPEG video stream is then retrieved from this
buffer by a software processing thread running on a host CPU. Every
time a picture start code is found in the video bit-stream buffer,
a full compressed MPEG picture, in the form of elementary stream,
is sent to the system memory buffer by DMA. One such method is
disclosed in U.S. patent application Ser. No. 09/489,682, entitled
"METHOD AND SYSTEM FOR ACCESSING PACKETIZED ELEMENTARY STREAM DATA"
(0100.9901350), which is hereby incorporated herein by
reference.
Before storing the full compressed MPEG picture in the DSM, a PES
packet header is added. The audio stream is de-multiplexed and
decoded by the host CPU. In a similar fashion as the video, prior
to audio decoding, the audio frames are packetized into PES
packets. Essential information from the PSI/SI/private data tables
is decoded and stored in a pure source form on a DSM. This way,
further reduction of the host DRAM memory requirements for queue
allocation and memory on the DSM media is reduced. An advantage of
this mode is reduction of CPU cycles needed for A/V playback of
stored data due to the PES format of audio/video data. PES
de-multiplexing is done in place, passing pointers to the payload
of PES packets that contain video or audio frames, other
implementations required they be sent by DMA to the video decoder
before they were decoded on host CPU (MPEG or AC-3 audio). As a
result, the host CPU doesn't move any raw audio or video data, and
host CPU utilization is reduced in order of magnitude compared to
TS playback operating mode.
In the foregoing specification, the invention has been described
with reference to specific embodiments. However, one of ordinary
skill in the art appreciates that various modifications and changes
can be made without departing from the scope of the present
invention as set forth in the claims below. For example, the
specific time-shifting implementation has been described as with
reference to a specific transport stream demultiplexer, and
described in a previous application which have been incorporated by
reference. Different transport stream demultiplexers and method of
implementing specific aspects of the present invention can be used
as well. Likewise, specific partitions between hardware and
software implementions have been described, which can vary
depending upon the implemented demultiplexer. For example, the
video stream parser can be designed to support routing the parsed
video data to a circular buffer that is accessible by the system
memory controller. Accordingly, the specification and figures are
to be regarded in an illustrative rather than a restrictive sense,
and all such modifications are intended to be included within the
scope of present invention. In the claims, means-plus-function
clause(s), if any, cover the structures described herein that
perform the recited function(s). The mean-plus-function clause(s)
also cover structural equivalents and equivalent structures that
perform the recited function(s). Benefits, other advantages, and
solutions to problems have been described above with regard to
specific embodiments. However, the benefits, advantages, solutions
to problems, and any element(s) that may cause any benefit,
advantage, or solution to occur or become more pronounced are not
to be construed as a critical, required, or essential feature or
element of any or all the claims.
* * * * *
References